Analysis

Two compounding query defects produced the false Berkhamsted zero, and the project's own registry already held the answer

The earlier finding that Great Berkhamsted held no Gurney burial at any date was produced by two query defects, each of which fails closed - returning zero rather than an error - and both generalise. First, a wildcard on the stem 'gurn' cannot reach 'Gourney', and the Berkhamsted family is indexed in the burial set predominantly under Gourney; the five seventeenth-century Gourney burials were invisible to the query that reported zero. Second, the place token differs between sets of the same county: the baptism set and the cross-collection view label the parish 'Great Berkhampstead', while the burial set labels it 'Berkhamstead', and because the keyword is a literal token match, both spellings tried earlier returned zero from a parish that is fully present. The project's own record already held the correction in two places - the source registry's note on the Hertfordshire burial set, recording a walk of 9 May 2026 that found Jhon Gourney burials of 1612 and 1620, and the identity assessment's own prose. A sweep contradicted both and reported the contradiction as a discovery. That is the mechanism the 'documented but never assembled' probability row describes, observed inside this project's own work.

Confidence: High

Topic file location: Identity — Berkhamsted: Candidates C(i) and C(ii)